The Law of Group Polarization

And an invitation to a group that operates according to a different law.

First things first: I wrote a Substack post on the mimetic process of cancel culture that went out to premium subscribers only on Sunday—just in case you missed me.

Please bear with me for 30 seconds while I give you a short elevator pitch on going premium—because it’s related to some exciting things in the works—and then we’ll get straight to the Girardian topic of this week’s newsletter: polarization.

Let me be clear: you’ll still get the kind of value I hope you’ve come to expect here as a free subscriber. In this edition, I’m going to share with you one of the most fascinating pieces of research related to mimetic theory that I’ve ever come across.

But as the publication of Wanting: The Power of Mimetic Desire in Everyday Life nears (June 1st, if you haven’t pre-ordered yet…), I’ll be transitioning a portion of these newsletters to premium subscribers only.

Anti-Mimetic was always meant to be “Wanting, Continued” in some respects—with the intention that it would lead to the formation of a community of thoughtful people who want to do meaningful work, and who recognize the need for developing anti-mimetic muscles so they don’t get sucked, tooth and nail, into the monoculture.

Thanks for being here. Now, I want to give you more ways to engage if you’d like to.

If you aren’t a premium subscriber, you’ll still receive the Anti-Mimetic newsletter…just a little less frequently.

If you do sign-up, you’ll receive full access to the archives plus some community features that are in the works:

  • A community chat platform (likely Discord) where we can have on-going conversations about the topics covered here. It’ll be like an AMA Forever. More importantly, it will allow the many wise people who read this newsletter to share their wisdom with everyone else.

  • Community Zoom calls around specific themes, and possibly IRL events—if there are enough people in a given city (I’ve noticed many of my American readers are clustered around a few major cities...)

  • I’m in the early stages of planning an “experience” (an un-Conference, if you will) on mimetic theory, sometime in 2022. Premium subscribers will have a first right of refusal to attend. I don’t see an invitation being extended to the general public at this point; we’ll want to keep it relatively intimate. But let’s see how much energy I have to plan that event after I finish planning my upcoming wedding… (If anyone is interested in being a part of this—the Mimetic Event, that is, not my wedding—please do reach out to me and the team.)

  • I’m working on an educational program and course on mimetic theory that serve as an intellectual introduction. In conjunction with that—and perhaps as a side journey to it—I’m developing an existential journey that is similar to the one I went on myself when I went into the desert. (If you were part of my Clubhouse event today, you got a taste of what that might be like.) Premium subs will have access to these educational and developmental tools at heavily discounted rates. (And by heavily, I mean heavy: $100 instead of $1,000—that sort of thing. The cost for a yearly subscription here is only $50.)

This newsletter is called The Anti-Mimetic Newsletter: A Field Guide to Mimetic Desire. I intend to start engaging much more closely with current events beginning shortly after the publication of Wanting.

The events unfolding will require us to reflect seriously on what’s happening—and to take appropriate action so that we can be people who desire with intention and help others to do the same.

(The most read/shared edition of this newsletter was my reaction to the Trump Twitter ban, in case you were wondering.)

Here’s what René Girard had to say:

“And there will be others, in any case, who will repeat what we are in the process of saying and who will advance matters beyond what we have been able to do. Yet books themselves will have no more than minor importance; the events within which such books emerge will be infinitely more eloquent than whatever we write and will establish truths we have difficulty describing and describe poorly, even in simple and banal instances.”

(That shouldn’t be interpreted as an excuse not to buy Wanting and to share it on social media with others to stoke their mimetic desire to do the same.)

Okay. Let’s finally get to the topic at hand: polarization.

Why “More Dialogue” Doesn’t Always Work

If you’re interested in why we’re so polarized, or highly dissatisfied with the Ezra Kleinan answer to that question, then I encourage you to download (free) this excellent paper by Cass R. Sunstein.

It’s one of the strongest pieces of research supporting how we know mimetic theory works in the ideological sphere that I have ever seen.

The paper, written in 1999 (for some odd reason, relatively unknown in Girardian circles), is called “The Law of Group Polarization.” Its core finding:

Contrary to popular belief, many people come away from group deliberation —including ”dialogue with people from the other side"—more polarized and extreme than they went in.

You can read the article yourself, so I will only summarize a few key points and draw attention to what I think is particularly salient to our exploration of mimetic theory and what it brings to bear on these issues.

First off, here’s Sustein giving his an explanation of what we know as mimetic desire, just using a different frame (that of “information theory,” as far as I can tell):

Social influences affect behavior via two different mechanisms. The first is information. As noted, what other people do, or say, carries an informational externality; if many other people go to a certain movie, or refuse to use recreational drugs, or carry guns, observers are given a signal about what it makes sense to do. The second mechanism is reputational. Even if people do not believe that what other people do provides information about what should be done, they may think that the actions of others provide information about what other people think should be done.

Now when it comes to group behavior, he cites the famous experiment by Solomon Asch in which people overrode their own sensory data when asked to select which of two lines on two separate sheets of paper were the same length.

A shocking percentage of people ignored their own sense of sight and tended to imitate others who were planted in the study to give wrong answers. This made people who had initially selected the correct line begin doubting themselves and even change their originally-correct selection to the wrong answer.

Under ordinary conditions, less than 1% of people would err. When introduced to the group pressure to select the wrong answer, subjects erred 36.8% of the time. In a series of twelve questions, over 70% of the subjects went along with the group. In other words, only 30% stuck with their original, correct answer.

Sustein calls this a social “cascade” (read: mimetic contagion). His words:

Some of the most interesting recent work on social influence [NB: forgive Sustein, he comes from academic that is hostile to Girard, so he has probably never heard of him] involves the possibility of informational and reputational “cascades”; this work has obvious relevance to law and politics. Indeed, it is possible to interpret Asch’s work as having demonstrated considerable individual susceptibility to cascade effects. What is striking about such effects is their ripple-like nature, or the quality of contagion. Group polarization is sometimes, but not always, a product of cascade effects.

From a Girardian perspective, group polarization is always a product of so-called “cascade” effects, which involves mimetic contagion that always seems perfectly rational to the people involved in it (and often times is rational—it’s rational to care about your reputation, for instance) but which leads people to highly mimetic clustering around certain issues. Oftentimes they cluster around an extreme.

A few (3) scant comments from me about this study:

  • What Sustein really seems to have found are two kinds of “dialogue.” One happens between people who enter into a dialogue without strong predispositions or presuppositions about the “Other side.” They enter into the deliberative process with a relatively open mind; at the very least, they don’t feel so strongly attached to an issue to the point that their identity is bound up in it at an existential level. They don’t view deliberation as an existential threat. These people can enter into a deliberative process with people who have different perspectives and emerge from the dialogue closer toward the proverbial middle (not that the “middle” is always a good place to be—it’s not; at least if you care about the truth), or at least not toward the extremes.

  • On the other hand, those who enter into these deliberative processes with strong predispositions and identity issues are essentially entering into a battle with a mimetic rival. There is no possibility of winning, or of moving closer to the middle. In this situation, we would come to expect both sides to move to extremes—it’s a battle of differentiation. This is especially relevant to our political discourse today. Most people do have strong predispositions toward their interlocutors as well as strongly held beliefs on which their identity rests—so it’s no surprise that Sustein found that the majority of people don’t benefit from the dialogue but rather become more polarized. I don’t think there’s anything too surprising about this. Lock members of the NRA in a room with people who want to abolish the 2nd amendment for an hour of forced conversation and nobody expects them to be singing Kumbaya after an hour; we would expect them to walk out of the room more antagonized and divided—or maybe worse.

  • An important part of this whole discussion, which Sustein doesn’t address in his research (because it’s not about mimetic theory explicitly), is that these two parties to the mimetic bind—the two parties that become polarized during the deliberative process—actually don’t become more “different” but more alike, at least on the level of desire. Their desire to vanquish the other mirrors that of their rival. Think of the two extremes on the ends of most of the hot-button political issues today and you will find more similarities than at first meets the eye—people who have become militant and who are willing to use violence to try to drive out violence. (Which, by the way, is one definition of the scapegoat mechanism.)

Give a gift subscription

An Anti-Mimetic Tactic

So what should we do?

In Sustein’s view, the “mechanisms that underlie group polarization raise serious questions about any general enthusiasm for deliberative processes. If the argument for deliberation is that it is likely to yield correct answers to social questions, group polarization suggests the need for attention to the background conditions in which this is likely to be the case.”

He admits that simply selecting a diverse group of people with “reasonable positions” to be part of a deliberative process runs into the circularity of self-selection.

The conceptual problem is that it’s impossible to specify what constitutes “appropriate heterogeneity” and the “appropriate plurality of views without making anecdote judgments about the substantive question at issue.”

This seems like a valid and pretty serious problem.

What is the anti-mimetic way, here?

I am not generally supportive of top-down approaches to engineering social situations in order to achieve desired outcomes. You typically solve one problem (if you’re lucky) while creating three more.

I do think there is some horizontal decision-rights delegation that can be done, though, which in my experience have yielded tremendously positive results.

Decision Rights Delegation

Horizontal delegation simply means delegating to peers decisions that I am perfectly capable of making myself but for which mimesis probably prevents me from achieving the most effective outcome.

We know how this works with non-mimetic decision-making. My friends can optimize for my health and safety in ways that I can’t. Here’s an example that everyone is familiar with: giving my car keys to a sober friend at the start of the night and letting him decide how we get home rather than assuming I’m going to be thinking clearly while I’m signing Sweet Caroline with the bartender at 2am (and I surely will believe that I’m thinking clearly—more clearly than I’ve ever thought, probably…).

Now simply transpose this to social situations in which mimetic desire is likely to be operative. Other people are not intoxicated from my own mimesis and are able to make certain decisions optimizing for my own intellectual growth—and other things—that I normally wouldn’t make for myself.

I enjoy holding Jeffersonian-style dinners (pre-pandemic, at least) in which I can select the people who will be there and where they sit in relation to other people. I think this prevents some of the natural mimetic movement and clustering that happens in these situations.

We all know how it works if you turn a high school class loose to “work in groups.” Certain students would never work with other students. When you first announce group work, they start looking around like deer caught in headlights to find their friends.

The same is true of us adults—but in far more subtle ways. And we usually don’t like to admit it.


I do believe that, in select circumstances, handing the power of association over to a trusted ringmaster—a connector—can break us out of the natural grooves of desire and dialogue in which we naturally want to move and in which our mimesis would keep us trapped were it not for an outside force.

I truly enjoy the work of thinking thoughtfully about connecting one person with another. While I haven’t played match-maker to any married couples yet, I have played match-maker to plenty of great friends, colleagues, and conversation partners.

So I propose this: one of the first “events” of the premium Anti-Mimetic community, once it launches in the next few months, will be (voluntarily, of course) an opportunity for me to personally match you up for a 30-minute, one-on-one conversation with another member of this community.

Obviously, I can’t do this until I’ve gotten to know each of you a bit. Part of that process will be a simple intake form that I’m in the process of designing based on some anti-mimetic design principles.

I know that it will lead to good conversations, at the very least.

At the very best, opportunities to collaborate together on important work: goods that are truly good and services that truly serve.

And that’s a positive kind of mimesis.